Vision Research 51 (2011) 1099–1108

Contents lists available at ScienceDirect

Vision Research

journal homepage: www.elsevier.com/locate/visres

A comparative study on the visual adaptations of four of moray ⇑ Feng Yu Wang a,b,1, Meng Yun Tang a,1, Hong Young Yan a, a Sensory Biology Laboratory, Marine Research Station, Institute of Cellular and Organismic Biology, Academia Sinica, Jiaoshi, I-Lan County 26242, Taiwan b Taiwan Ocean Research Institute, Taipei 10622, Taiwan article info abstract

Article history: The goal of this study was to investigate how the eyes of different species of evolved to cope Received 18 November 2010 with limitations to vision imposed on them by the photic environments in which they reside. The com- Received in revised form 22 February 2011 parative retinal histological structures and visual pigment characteristics including opsin gene sequences, Available online 6 March 2011 of four species of moray eel inhabiting diverse habitats (i.e., shallow-water species, Rhinomuraena quae- sita and favagineus, and deep-sea species, Gymnothorax reticularis and Strophidon sathete) Keywords: were examined. The histological sections showed that retinal layer structures of R. quaestia are signifi- Moray eel cantly different from those of the other three species which likely reflects the effects of distribution depth Microspectrophotometry (MSP) on the structures. The maximal absorbance wavelength (kmax) of photoreceptor cells, as measured by kmax Visual characteristics microspectrophotometry (MSP), showed a close correlation between the kmax and the intensity/spectral Opsin gene quality of the light environment where each species lives. The spectra-shift, between shallow and deep- sea species, observed in the rods cells results from amino acid substitution in Rh1 gene, while that in cones most likely results from differential expression of multiple Rh2 genes. Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction Vision begins when photons are absorbed by photoreceptors in the retina. Two types of photoreceptors are found in most vertebrate The solar irradiance measured at depth in natural waters is retinas – rods and cones. Rods mediate scotopic vision and generally influenced by the absorptive characteristics of the water as well have long, cylindrical outer segments. Cones mediate photopic, high as the time of the day, suspended particle, nutrient load, phyto- acuity vision, and usually have shorter, more conical outer seg- plankton and zooplankton concentrations. Due to these factors, ments. They can exist as single cells or into coupled groups as dou- the photic environment of aquatic organisms exhibits a great bles or even triples (Sandström, 1999). Both types of photoreceptors diversity of irradiant and optical conditions. In order to adapt to contain visual pigments, which are composed of an opsin protein the wide extent of specific photic environments, such as those and a chromophoric group, either 11-cis-retinal (based on vitamin found in estuaries, coastal, shallow, deep-sea, rivers and lakes, A1) or 11-cis-3-dehydroretinal (based on vitamin A2). In vertebrates, fishes have evolved various visual system characteristics allowing there are five opsin gene families giving rise to the visual pigments them to operate under different types of photic conditions (Loew (Yokoyama, 1994, 1995, 1997; see Bowmaker & Loew, 2008). Rh1 is

& McFarland, 1990). As solar radiation penetrates clear blue oce- expressed in the rods and yields vitamin A1-based visual pigments anic water, the shorter wavelengths (i.e., blue light; ca. 400– having kmax from 460 to 530 nm (Yokoyama, 1997). The vitamin 500 nm) are absorbed less than longer wavelengths resulting in a A1-based visual pigments found in cones formed by the other four narrowing of the visible spectrum at depth with the peak of the expressed opsin genes are a long- to middle-wave class (LWS) max- downwelling light being in the region of 435 nm (Kirk, 1983). In imally sensitive in the red–green spectral region from about 490– coastal and fresh water the increase in dissolved organics, i.e., 570 nm, a middle-wave class (RH2) sensitive in the green from the so-called ‘‘Gelbstoff’’ and scattering particulates shifts the about 480–535 nm, a short-wave class (SWS2) sensitive in the transmission maximum to longer wavelengths (Jerlov, 1968). blue–violet from about 410–490 nm and a second short-wave class Therefore, in clear water, the photic environment exists as a (SWS1) sensitive in the violet–ultraviolet from about 355–440 nm blue–green color, while the spectrum of the ambient light in coast- (Bowmaker, 2008; Bowmaker & Loew, 2008; Bowmaker, Semo, al and lake waters would be more in the green to orange wave- Hunt, & Jeffery, 2008; Ebrey & Koutalos, 2001; Yokoyama, 2000; length range (McFarland, 1986; Morel, 1980). Yokoyama & Yokoyama, 1996). A number of visual system adaptations allow fish to cope with

⇑ Corresponding author. Fax: +886 3 9871035. the constraints imposed by a habitat’s specific photic environment. E-mail address: [email protected] (H.Y. Yan). First, variations in eye and retinal structure allow some fishes to 1 These authors contributed equally. exploit different habitats and niches more effectively (Bowmaker,

0042-6989/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.visres.2011.02.025 1100 F.Y. Wang et al. / Vision Research 51 (2011) 1099–1108

1990, 1995; Collin, 1997). For example, fishes that live in deep-sea 2002); (2) the deep-water group, consisting of the dusky-banded environments have adaptations that address the problems of low moray, Gymnothorax reticularis (depth range: 30–200 m) and the light intensity such as larger eyes or a tapetum which reflects light slender , Strophidon sathete (depth range: 1–300 m), back (Nicol & Somiya, 1989; Warrant & Locket, 2004). There may which live in sand–muddy sediment (Randall, Allen, & Steene, also be longer outer segments that increase the probability of pho- 1990; Smith & Bohlke, 1997). Since the habitats of these two ton capture or banked retinas (see McFarland, 1991). The problems groups of moray differ so much in their respective photic envi- of the spectral shifts in background space light due to depth and ronments, comparisons of the differences between these two changes in water quality have been addressed by altering the groups could provide useful information to delineate how moray absorptive properties of the visual pigments either by amino acid eels evolved to cope with the environmental constraints in terms alterations of visual pigment opsins that create visual pigments of light conditions. more appropriately ‘tuned’ to the visual tasks present, or by alter- In this study, histological methods were used to measure the ing the expression pattern of the opsin genes, or both (Bowmaker thickness of each retinal layer with the expectation that increases et al., 2008; Carleton & Kocher, 2001; Cottrill et al., 2009; Parry in photoreceptor and outer nuclear layer thicknesses would be et al., 2005; Shand, Hart, Thomas, & Partridge, 2002; Shand et al., associated with the dim light condition. Second, the absorption 2008). There is also the possibility of switching chromophore class spectra of the photoreceptor cells were obtained by microspectro-

(vitamin A1- to vitamin A2-based) or employing some kind of pho- photometry (MSP). Finally, the opsin genes from these four moray tosensitizer as has been found for some deep-sea species (see Bow- eel species were cloned and sequenced. The combination of these maker & Loew, 2008). data allow us to speculate on how moray eels have adapted to their Numerous studies have documented the changes associated photic environments. with the retinas and visual pigments of fishes inhabiting different photic environments. For visual pigments, the findings have been 2. Materials and methods interpreted in the context of two hypotheses. The Sensitivity Hypothesis states that for maximizing the brightness contrast of 2.1. Samples collection a target against its background a single photoreceptor visual pig- ment k should be located close to the maximum of the down- max The moray eel species used in this study were obtained in a welling space light to maximize quantum catch. Thus, the k of max variety of ways. Specimens of R. quaesita (ribbon eel) were im- rod visual pigments shifts to shorter wavelengths as habitat depth ported from Southeast Asian waters via a vendor in Singapore. G. increases (see Bowmaker, 2008). The Contrast Hypothesis states favagineus (laced moray) were bought in Bi-Sha Fishing Harbor, that two visual pigments are necessary for maximizing chromatic Keelung, Taiwan, where they were caught with plastic tubing traps (i.e., color) contrast – one with its absorbance matched to the back- at a depth of approximately 30 m around Peng-Hu Archipelagos, in ground space light and the other offset from the background so as the middle of Taiwan Strait. G. reticularis (dusky-banded moray) to maximize the difference in the background and target chroma- and S. sathete (slender giant moray) were caught by bottom trawl- ticities (see Bowmaker, 2008). ers from depths of 50–800 m and landed in Da-Si Fishery Harbor, I- Numerous fish groups from different habitats have been exam- Lan, Taiwan. All specimens were kept in a tank with running sea- ined for their visual pigment complement and their retinal struc- water (temperature of 25–28 °C) under natural light cycle at the ture. However, few have been conducted on members of eel Marine Research Station, Institute of Cellular and Organismic Biol- families, including freshwater eels and the conger eels. To adapt ogy, Academia Sinica, Taiwan. They were fed with fish meat ad libi- to the deep-sea environments, freshwater eels (Anguilla spp.) and tum three times a week until use. The use protocols used in conger eels (Conger spp.) possess photoreceptors with a blue- this study were approved by Academia Sinica Institutional Animal shifted kmax (Archer & Hirano, 1996; Denton & Walker, 1958; Shap- Care and Use Committee (No. RFiZOOYH2007012). ley & Gordon, 1980). Moreover, freshwater eels can alter their spectral sensitivities during their migration from the freshwater to the deep-sea environment either by switching chromophore 2.2. Histology and samples preparation type, or by expressing different opsin genes to cope with the changing light environments (Bowmaker et al., 2008; Cottrill All specimens were dark-adapted overnight (at least 6 h) inside et al., 2009). a darkroom prior to use. Under infrared light illumination, with the Moray eels are generally recognized as nocturnal predators be- aid of a pairs of night vision goggle (Bushnell-Night Eye M220) and cause of their relatively smaller eyes and well-developed olfactory a dissecting stereomicroscope, the fishes were first anesthetized sense and sensory pores, all of which could enhance their foraging with MS-222 (50 ppm), and then the eyes were enucleated. The ability during the night (Bardach & Loewenthal, 1961; Bardach, cornea, lens and vitreous humour were removed from both eyes Winn, & Menzel, 1959; Hess, Melzer, & Smola, 1998; Winn & Bard- of each fish. The retina of one eyecup, intended for MSP measure- ach, 1959; Young & Win, 2003). However, some moray eel species ment, was separated from the pigment epithelium and immedi- have been reported to forage during the day relying on their eyes ately immersed in chilled phosphate buffered saline (Sigma, USA; (Böhlke & Randall, 2000; Chave & Randall, 1971; Hobson, 1975). pH 6.5); the other eyecup, used for histological study, was fixed This contradictory information seems to imply that moray eel spe- in Bouin’s solution (Ricca Chemical Company, No. 1120-32). cies may have different visual perceptual abilities in terms of re- For histological analysis, retina preparations were then dehy- sponses to light, i.e., color perception. drated through a series of ethanol solutions, embedded in paraffin, Four species of moray eels in the subfamily Muraeninae were sectioned at 5 lm, and stained with hematoxylin and eosin (H&E). selected to conduct a comparative study on their retinal structure Radial sections of the retina were examined under a light micro- and their visual pigment/opsin gene complement. In terms of the scope. In order to compare the differences of overall structures of depth of environments where they reside, these four species can the retinae among the four species, retinal preparations from two be divided into two groups: (1) The shallow-water group, consist- adult individuals of each studied species were used. The thick- ing of the ribbon eel, Rhinomuraena quaesita (depth range: 1–57 m) nesses of four distinct layers, including pigment epithelium (PE) and the laced moray, Gymnothorax favagineus (depth range: 1– layer, photoreceptors layer (PL, layer of rod and cone cells), outer 45 m). These two species are crevice-dwelling predators inhabiting nuclear layer (ONL, layer of nuclei of photoreceptors), and inner coral reefs in shallow seas (Böhlke & Randall, 2000; King & Fraser, nuclear layer (INL, layer of cell body of interneurons) were F.Y. Wang et al. / Vision Research 51 (2011) 1099–1108 1101 measured from the retinal sections using a calibrated ocular Alignments of the Rh1 and Rh2 genes were carried out using micrometer. A Kruskal–Wallis one-way analysis of variance on their predicted amino acid sequences with CLUSTAL W function in ranks (ANOVA on Ranks) complemented by post hoc Dunn’s multi- the MEGA 3.1 software (Kumar, Tamura, & Nei, 2004), and nucleo- ple comparisons test were used to compare the differences in tide sequences were aligned in accordance with the amino acid thickness among distinct layers of retina in the four moray eel alignments. The best-fit model of nucleotide evolution was deter- species. mined by hierarchical likelihood ratio tests (LRT) implemented in Model Test v3.7 (Posada & Crandall, 1998). The PAUPÃ 4.0 (Swofford, 2.3. Microspectrophotometry (MSP) 2000) was used to construct neighbor joining phylogenetic trees (Saitou & Nei, 1987) by applying ML distances from the best-fit Microspectrophotometry was carried out on 4–11 adult individ- model and 1000 bootstrap replicates. Ancestral sequences of the op- uals of each studied species. Absorbance spectra of individual sin genes of each moray eel species was estimated by using PAML photoreceptors were measured using a computer-controlled, (Yang, 1997, 2007). single-beam microspectrophotometer, which has been previously described (see Loew, 1994) and used in our previous studies 3. Results (Wang, Chung, Yan, & Tzeng, 2008; Wang, Yan, Chen, Wang, & Wang, 2009). The retina was cut into small pieces, placed on a cov- 3.1. Retinal morphology er glass, macerated and covered with a smaller glass cover slip edged with silicone grease. The preparation was then placed onto The paraffin radial sections with H&E staining revealed that the the microspectrophotometer stage. A baseline absorbance spec- moray eel retina was a duplex retina, including rods and one mor- trum was obtained from a cell-free area of the preparation, fol- phological type of cone cell. This is consistent with the report of Ali lowed by the absorbance spectrum from the outer segment. The and Anctil (1976). While the overall picture for the four studied kmax of the measured visual pigment was obtained by a pro- moray eel species revealed a commonality of basic structure, there grammed statistical method (Loew, 1994). The methods used to were some differences in terms of thickness of each distinct layer estimate the kmax and the A1/A2 template of the normalized absor- among species. As shown in Fig. 1, the retina can be classified into bance spectrum followed those previously used protocols (Gov- four different layers: pigment epithelium (PE) layer, photorecep- ardovskii, Fyhrquist, Reuter, Kuzmin, & Donner, 2000; Lipetz & tors layer (PL, layer of rod and cone cells), outer nuclear layer Cronin, 1988). The determination of the best-fit template was (ONL, layer of nuclei of photoreceptors), and inner nuclear layer made by visual examination with the lowest standard deviation (INL, layer of cell body of interneurons). The thickness distributions (SD). If the SD of kmax was smaller than 7.5 nm, then the spectrum of each layer for each species are shown in Table 1. R. quaesita is was considered valid and stored in the computer (Loew & Sillman, unique because its PL and ONL layers were the thinnest among 1993; Sillman, Johnson, & Loew, 2001). This process was repeated the four species, while its INL thickness was the greatest (Fig. 1). for each photoreceptor examined by the MSP. After the kmax values Its mean PE thickness was not significantly different from that of of each photoreceptor were averaged, a final estimate of mean G. favagineus, but was significantly thicker than those of G. reticu- kmax ± SD was obtained. A t-test was used to compare the differ- laris and S. sathete. Further, R. quaesita had the lowest ONL/INL ratio ences among the spectral sensitivities of these four species. (Table 1). Based on the analysis of ANOVA on ranks (data not shown), the 2.4. cDNA synthesis, primers design and PCR amplification of opsin examined moray eel species could be divided into two groups. genes Group 1, including G. favagineus, G. reticularis and S. sathete, had a thin pigmented epithelial layer. Their rod cells were slender, Total RNA was extracted from freshly dissected retinae using a elongated, and numerous. Additionally, their cone cells, with short QIAGEN RNeasy Mini kit (Valencia, California, USA). Single- outer segments, were very small. These species possessed an extre- stranded cDNA was synthesized using an oligo-d(T) primer and mely thick and well-development ONL, however their INL was rel- SuperScript II Reverse Transcriptase (Invitrogen, Carlsbad, Califor- atively thin. Group 2 includes only one species: R. quaesita in which nia, USA). The design of primers used for opsin gene amplification the epithelial cell processes were well developed and were filled was based on the conserved regions of the opsin genes from cypri- with melanin pigments. The cone cells were present in large num- nids, sparids, freshwater eels, and cichlids (Carleton & Kocher, bers, whereas the rod cells were fewer and rather short. Further- 2001; Cottrill et al., 2009; Spady et al., 2005; Wang et al., 2008, more, the ONL was poorly developed and was thinner than the 2009). This primer list is available in Supplementary Table S1. INL (Fig. 2). The PCR reaction solution contained 1 ll cDNA, 4 ll Fast Run™ Â5 Taq master mix (Protech, Taipei, Taiwan), with 0.5 ll (5 mM) of each primer added to 20 ll of double distilled H2O. Reactions 3.2. Visual pigments were conducted on a Thermal Cycler (Veriti Thermal Cycler of Ap- plied Biosystems, Foster City, California, USA) at 95 °C for 30 s, Data in Table 2 showed the mean values of kmax for rods of the 50 °C for 30 s, 72 °C for 1 min for 35 cycles, and a final extension four species examined: R. quaesita (498 ± 4.8 nm), G. favagineus at 72 °C for 5 min. (487 ± 5.4 nm), G. reticularis (486 ± 4.0 nm), and S. sathete

(487 ± 4.8 nm). The t-tests showed that the kmax among G. favagin- 2.5. Cloning, sequencing and sequence analysis eus, G. reticularis and S. sathete are not significantly different

(P > 0.05); however, the kmax of R. quaesita was significantly differ- The PCR products of the five cone opsins were cloned individu- ent from the other three species (Table 3). The kmax of G. favagineus, ally into T-vectors using pGEM@-T vector system (Promega, G. reticularis and S. sathete showed an 11–12 nm blue-shift when Madison, Wisconsin, USA), and then sequenced respectively. Com- compared to that of R. quaesita (Table 1). In conjunction with the mercial sequence kits (BigDye™ Terminator Cycle Sequencing known distribution depth of the four moray eel species, these data

Ready Reaction Kits of Applied Biosystems, Foster City, California, indicate that the kmax of rod cells exhibited a blue-shifted pattern USA) and ABI model 377 automated DNA sequencers were used with increasing habitat depth. to obtain sequence data. Five to ten clones of each opsin gene were The four moray species examined all possessed only one single sequenced to rule out the artificial errors. cone spectral class. Table 1 shows that the mean kmax values for cone 1102 F.Y. Wang et al. / Vision Research 51 (2011) 1099–1108

A: R. quaesita B: G. favagineus C: G. reticularis D: S. sathete

1 1 1

1 2 2 2 2 3 3 3 4 3

4 4 4

10 µm

Fig. 1. Photomicrographs of transverse histological sections of retina of the four moray eel species examined. (1) PE: pigment epithelium layer of the retina; (2) PL: photoreceptors layer, i.e., layer of rod and cone cells; (3) ONL: outer nuclear layer; layer of nuclei of photoreceptors; (4) INL: inner nuclear layer; layer of cell bodies of interneurons. Scale bars = 10 lm.

Table 1 The mean thicknesses of four distinct layers of retinas among the four moray eel species examined. Data were presented as mean thickness ± SD (lm). N: indicates the number of specimens examined. n: indicates the number of histological sections counted. PE: pigment epithelium of retina; PL: photoreceptors layer, i.e., layer of rod and cone cells; ONL: layer of nuclei of photoreceptor; INL: layer of cell body of interneurons.

Species Thickness (lm) ± SD Ratio of ONL to INL PE PL ONL INL Rhinomuraena quaesita N = 2 n = 20 4.34 ± 0.47 4.31 ± 0.53 3.88 ± 0.72 8.34 ± 0.79 0.47 Gymnothorax favagineus N = 2 n = 14 3.62 ± 0.86 8.26 ± 1.55 9.78 ± 0.72 4.20 ± 0.74 2.39 Gymnothorax reticularis N = 2 n = 14 2.14 ± 0.64 10.63 ± 1.15 8.66 ± 0.69 4.33 ± 0.57 2.05 Strophidon sathete N = 2 n = 22 2.76 ± 0.37 9.69 ± 1.15 7.16 ± 0.69 5.20 ± 1.08 1.43

A. R. quaesita-Rod C. G. favagineus-Rod E. G. reticularis-Rod G. S. sathete-Rod 0.05 0.035 0.10 0.07 0.030 0.06 0.04 0.08 0.025 0.05 0.03 0.06 0.020 0.04 0.02 0.015 0.04 0.03 Absorbance Absorbance Absorbance Absorbance 0.010 0.02 0.01 0.02 0.005 0.01 0.00 0.000 0.00 0.00

300 400 500 600 700 800 300 400 500 600 700 800 300 400 500 600 700 800 300 400 500 600 700 800 Wavelength (nm) Wavelength (nm) Wavelength (nm) Wavelength (nm)

B. R. quaesita-Cone D. G. favagineus-Cone F. G. reticularis-Cone H. S. sathete-Cone 0.05 0.04 0.06 0.05 0.05 0.04 0.04 0.03 0.04 0.03 0.03 0.02 0.03 0.02 0.02 0.02 Absorbance Absorbance Absorbance Absorbance 0.01 0.01 0.01 0.01 0.00 0.00 0.00 0.00

300 400 500 600 700 800 300 400 500 600 700 800 300 400 500 600 700 800 300 400 500 600 700 800 Wavelength (nm) Wavelength (nm) Wavelength (nm) Wavelength (nm)

Fig. 2. The representative absorbance spectra of the rod (upper row) and cone (lower row) of the four moray eel species examined. Black lines: absorbance spectra of photoreceptor cells before bleaching; grey lines: absorbance spectra of photoreceptor cells after bleaching.

cells were: R. quaesita (493 ± 7.0 nm), G. favagineus (501 ± The t-test showed that the kmax between G. reticularis and R. quaesita 7.7 nm), G. reticularis (494 ± 5.8 nm) and S. sathete (509 ± 6.6 nm). were not significantly different (P > 0.05) (Table 2). The representa- F.Y. Wang et al. / Vision Research 51 (2011) 1099–1108 1103

Table 2 3.4. Opsin phylogeny The mean kmax of rod and cone cells from moray eels measured by MSP. All values are expressed in nanometers (nm) with mean ± SD. The Rh1 genes of conger eel, fresh eels, pacific salmonids and Species Rod Cone cells (green single cichlids; and the Rh2 genes of cichlids were included in the phylo- cone) genetic analysis, while those of goldfish, carp and zebrafish were Rhinomuraena quaesita 498 ± 4.8 nm 493 ± 7.0 nm (n = 37) used as out-groups (Fig. 4). A neighbor-joining tree of opsin genes (N = 11) (n = 101) was constructed based on the best-fit model by applying ML dis- Gymnothorax favagineus 487 ± 5.4 nm 501 ± 7.7 nm (n = 47) tances and 1000 bootstrap replicates. The Rh1 gene of moray eels (N =5) (n = 81) Gymnothorax reticularis 486 ± 4.0 nm 494 ± 5.8 nm (n = 40) clustered together to form a monophyletic group, which is the sis- (N =8) (n = 60) ter group of the freshwater type Rh1 gene of conger and freshwater Strophidon sathete (N = 4) 487 ± 4.8 nm 509 ± 6.6 nm (n = 29) eels (Fig. 4). The Rh2 genes of moray eels and freshwater eel (n = 76) formed a monophyletic group. Rh2A genes of moray eels clustered N: indicates the number of the specimens examined. n: indicates the number of together and formed a sister group of Rh2B in moray eels. photoreceptor cells measured.

4. Discussion Table 3 The results of t-test of spectral sensitivities among rod and cone cells of studied 4.1. Retinal morphology of moray eels differed between diurnal and moray eels. The upper rows (white) present the t-test results of rod cells while the lower rows (grey) present those of cone cells. nocturnal species

Retinal structure have been shown to reveal unique features be- tween diurnal and nocturnal teleosts (Ali & Anctil, 1976; Munz & McFarland, 1973; Pankhurst, 1989; Walls, 1942). In nocturnal tele- osts, the PE and INL were usually thinner, but the PL and ONL were relatively thicker. However, the diurnal teleosts have exactly the ÃÃIndicates P < 0.001; NS, no significant difference: P > 0.05. opposite metrics. The retinal sections of Group 1, including G. fava- gineus, G. reticularis and S. sathete, showcased the characteristics of tive examples of absorbance spectra of rod and cone cells were pre- nocturnal species. R. quaesita, of Group 2, possessed retina histolog- ical features typical of a diurnal species. Moreover, the length ratio sented in Fig. 2, while the frequency distributions of kmax of both rod and cone cells of each species were presented in Fig. 3. of ONL to INL could also be used as an indicator of nocturnal or diur- nal species: fishes with greater ONL/INL ratio are regarded as a noc- turnal species, while a smaller ratio indicates a diurnal species 3.3. Opsin gene sequences of moray eels and amino acid substitutions (Munz & McFarland, 1973). The ONL/INL ratio of G. favagineus at tuning sites (2.39) and G. reticularis (2.05) indicated that they were nocturnal species, while that of R. quaesita (0.47) indicated it as a diurnal spe- Partial opsin genes were amplified and cloned from cDNA of the cies. S. sathete, which had a ratio value of 1.43, would be considered retinae of moray eels. Only two opsin genes, Rh1 and Rh2, were as crepuscular or diurnal species (Munz & McFarland, 1973). These found in the four moray eel species, while SWS1, SWS2 and LWS retinal structure data revealed that not all moray eels should be re- gene failed to be amplified by using degenerated primers. This garded as nocturnal species as previously thought. finding is consistent with our MSP data, i.e., moray eels are green-light sensitive only. The size of opsin gene amplified was 931 bps for Rh1 (from amino acid site 32–341) and 567 bps for 4.2. The spectral position of rod visual pigments supports the the Rh2 gene (from amino acid site 117–306), respectively. sensitivity hypothesis In the Rh1 gene, there are seven amino acid sites important for spectral tuning: 83, 122, 211, 261, 265, 292, and 295 (Yokoyama, Spectral sensitivities of rods, assumed from the visual pigment 2000). The seven sites were conserved across moray eels, except absorbance spectra, in the different moray eel species examined at site 292 in R. quaesita (Table 4). In R. quaesita, there was a sub- correlated well with the depth of their habitats, except for G. fava- stitution of S292A (change from serine to alanine), relative to the gineus, which lives in shallow seas but has deep-sea type rhodopsin. consensus sequence (Table 4). Earlier studies, based on site-direc- In order to adapt to the photic conditions of the deeper ocean, the ted mutagenesis, suggest that a substitution of S292A and A292S kmax of rod cells in G. reticularis and S. sathete exhibited 12 nm could induce 7–16 nm red-shift and 7–15 nm blue-shift of kmax, blue-shifted to conform with the light spectra of their habitats. respectively (Archer, Hope, & Partridge, 1995; Davies et al., 2009; Numerous investigators have established that as light penetrates Fasick & Robinson, 1998; Hunt, Fitzgibbon, Slobodyyanyuk, & water, the shorter wavelengths representing blue light are trans- Bowmaker, 1996; Takenaka & Yokoyama, 2007; Yokoyama, 2008; mitted more readily than those of longer wavelengths (Loew & Yokoyama, Tada, Zhang, & Britt, 2008). This observation is consis- McFarland, 1990; Loew & Zhang, 2006). The spectral sensitivity of tent with our MSP data. rod visual pigments of coral reef fishes tend to be more blue-shifted Two copies of the Rh2 opsin gene, Rh2A and Rh2B, were found than in deeper water species (Cummings & Partridge, 2001; Losey in G. favagineus, G. reticularis and R. quaesita, while S. sathete pos- et al., 2003). A similar relationship between kmax and living depth sessed only one copy of Rh2A (Table 4). Amino acid substitutions was also observed in the present study (Fig. 3). The kmax of rods of at sites 97, 122, 207 and 292 could result in the spectral shift of R. quaesita, 498 nm, was suitable for shallow-water photic condi- the Rh2 gene (Takenaka & Yokoyama, 2007). Site 97 was not in- tions. To adapt to the photic conditions of the deeper ocean, G. retic- volved in our PCR amplification. In Rh2A and Rh2B genes, no differ- ularis and S. sathete have shifted their kmax of rods to 486 nm to ences were found among the tuning sites in moray eels (Table 4). better match the bluer photic environment of their habitats. E122Q (glutamic acid to glutamine) substitution, which could in- It is known that G. favagineus lives at a depth of approximately duce blue-shift in Rh2 opsin pigments (Takenaka & Yokoyama, 45 m; however, its rod kmax at 487 nm, was similar to those of the 2007; Yokoyama, 2000, 2008), was found between Rh2A and deep-sea species. This inconsistency could result from three possi- Rh2B of moray eels (Table 4). bilities. First, G. favagineus could be a species which migrates 1104 F.Y. Wang et al. / Vision Research 51 (2011) 1099–1108

A. R. quaesita B. G. favagineus

40 40 Cone cell Cone cell Rod cell Rod cell

30 30

20 20

10 10

0 0 471-475 476-480 481-485 486-490 491-495 496-500 501-505 506-510 511-515 516-520 471-475 476-480 481-485 486-490 491-495 496-500 501-505 506-510 511-515 516-520

C. G. reticularis D. S. sathete

40 40 Cone cell Cone cell Rod cell Rod cell Numbers of photoreceptor cells Numbers of photoreceptor

30 30

20 20

10 10

0 0 471-475 476-480 481-485 486-490 491-495 496-500 501-505 506-510 511-515 516-520 471-475 476-480 481-485 486-490 491-495 496-500 501-505 506-510 511-515 516-520 Wavelength (nm)

Fig. 3. Pooled data of distribution histograms of maximal absorbance wavelengths (kmax) of photoreceptor cells examined in four moray eel species. (A) R. quaesita; (B) G. favagineus; (C) G. reticularis; (D) S. sathete. Rod cells: hatched bars. Cone cells: open bars.

Table 4 Comparisons of the opsin sequences of moray eels.

Sequences are compared to the consensus sequences with similar identity indicated by a dot. The dash bars indicate that sequences are not available in this study. Sites are numbered according to bovine rhodopsin. kmax from MSP (in nm) are listed for those genes that are expressed in moray eels. The white and grey rows indicate the moray eels that inhabit shallow and deep-sea, respectively. vertically between shallow and deeper waters. For example, the 1980). Second, the rods could be adaptive for twilight vision. It is conger eel, which migrates vertically daily possesses rod cells with known that the overall twilight spectrum is weighted towards a blue-shifted kmax (487 nm) to adapt to both dim light conditions the blue by atmospheric absorption at low solar angles. This is while in the deep-sea during the day and shallow water photic the ‘Twilight Hypothesis’ of MacFarlane and Munz (McFarland, environments at night (Archer & Hirano, 1996; Shapley & Gordon, 1986; McFarland & Munz, 1975; Sandström, 1999). The kmax of F.Y. Wang et al. / Vision Research 51 (2011) 1099–1108 1105

Fig. 4. Neighbor joining trees of the moray eel Rh1 (A), Rh2 (B) opsin genes based on the ML distances from the best-fit model of model test. The model HKY + G (Hasegawa, Kishino, & Yano, 1985; Posada & Crandall, 1998) was used for constructing the phylogenetic trees of the Rh1 genes, and model TrN + I + G (Posada & Crandall, 1998; Tamura & Nei, 1993) for the Rh2 gene. A292S, S292S and E122Q in the trees showed the substitution events occurred, and the grey marks indicated the gene duplication events. The length of scar bar indicated the 0.05 nucleotide substitution. The GenBank accession numbers of the opsin genes of moray eels in this study were listed as following, Rhinomuraena quaesita: Rh1 HQ444180, Rh2A HQ444184 & Rh2B HQ444185; Gymnothorax favagineus: Rh1 HQ444181, Rh2A HQ444186 & Rh2B HQ444187; Gymnothorax reticularis: Rh1 HQ444182, Rh2A HQ444188 & Rh2B HQ444189; Strophidon sathete: Rh1 HQ444183 & Rh2A HQ444190. The nucleotide sequences of fish opsin genes were obtained from GenBank: Zebrafish Rh1 (BC164171), Rh2–1–2–4(AB087805, AB087806, AB087807, AB087808); Goldfish Rh1 (L11863), Rh2–1 and Rh2–2 (L11865, L11866); common carp Rh1(Z71999); Dimidiochromis compressiceps Rh1 (AY775059); Oncorhynchus nerka Rh1 (AY214156); Oncorhynchus keta Rh1 (AY214141); Anguilla anguilla Rh1 freshwater and deep-sea type (AJ249202, AJ249203); Anguilla japonica Rh1 freshwater and deep-sea type (AJ249202, AJ249203); Conger myriaster Rh1 freshwater and deep- sea type (AB043817, AB043818); Melanochromis vermivorus Rh2Aa, Rh2Ab and Rh2B (DQ088631, DQ088634, DQ088646); Pseudotropheus acei Rh2Aa, Rh2Ab and Rh2B (DQ088630, DQ088633, DQ088645); Anguilla anguilla Rh2 (FJ515778). 1106 F.Y. Wang et al. / Vision Research 51 (2011) 1099–1108 rod visual pigments of moray eels were in agreement with those Yokoyama et al., 2008). When polar amino acids, such as Ser and known rod kmax values which could optimize the photon absorp- Thr, occur at this site of Rh1 opsin, the kmax was usually around tion during dawn and dusk (Table 2)(McFarland & Munz, 1975; 485 nm (Hunt et al., 2001). Substitution from alanine to serine at Munz & McFarland, 1973). Finally, a possibility, which cannot be site 292 (A292S) could induce a 7–15 nm blue-shift, but substitu- ruled out, is that G. favagineus is just a shallow-water species but tion from serine to alanine could have the opposite effect with a possess a deep-sea type rod cells. This phenomenon is not rare. 7–16 nm spectral shift toward red (Archer et al., 1995; Davies

For example, the parrotfish (Scaridae) possess a blue-shifted kmax et al., 2009; Fasick & Robinson, 1998; Hunt et al., 1996; Takenaka of rod cells at 483–485 nm, which was deemed inconsistent with & Yokoyama, 2007; Yokoyama, 2008; Yokoyama et al., 2008). Our their shallow-water photic environment. However, the blue- Rh1 gene data was consistent with these findings. The substitution, shifted of kmax could reduce the photoreceptor noise to enhance S292A, was observed in the Rh1 opsin of R. quaesita relative to that performance at low light environments (Bowmaker, 1995), and of its deep-sea counterparts, G. reticularis and S. sathete, and could could provide the fish with higher visual sensitivity at greater result in a 12 nm red-shift to adapt to the photic environment of depths, in order to detect potential predators during twilight shallow-sea environments. Therefore, these four moray eel species migration (Munz & McFarland, 1973; Ogden & Buckman, 1973). appear to have used substitution at spectra-tuning site 292 of Rh1

The kmax of cone cells among the studied moray eels, however, gene to produce the rod spectral shift. seems not to correspond well with differences in habitat depth. Amino acid substitution at site 122 is an important site for spec- The known diurnal species, R. quaesita, is active during daytime. tral tuning of Rh2 genes (Takenaka & Yokoyama, 2007; Yokoyama,

However, its kmax of cone cells at 493 nm, also suggests that R. quae- 2000, 2008; Yokoyama, Zhang, Radlwimmer, & Blow, 1999). Sub- sita could be suited for twilight vision. This could allow for an in- stitution from glutamate to glutamine at site 122 (E122Q) can in- crease in their activity and vision under dawn and dusk. duce a blue-shift of the Rh2 gene in fishes (Chinen, Matsumoto,

Furthermore, the kmax of cones of G. reticularis, 494 nm, could con- & Kawamura, 2005; Wang et al., 2008; Yokoyama, 2008; Yokoyama form to the dim light photic conditions at the depth around 200 m. et al., 1999). In Rh2 genes of cichlids and zebrafish, the kmax was The penetration of light in coastal waters would be slightly red- shorter than 500 nm when glutamine was present at site 122, shifted, which is caused by suspended particles and dissolved organ- while kmax was longer than 500 nm when glutamic acid was at this ic materials including planktons. Therefore, the green-light, ca 500– site (Chinen et al., 2005; Parry et al., 2005). In moray eels, the Rh2A 530 nm, could penetrate into the deeper depth (McFarland, 1986). had glutamic acid at site 122, and on the contrary, Rh2B had gluta-

The kmax of cone cells of G. favagineus, 501 nm, displayed an 8 nm mine at the same site. Moreover, numerous studies in cichlids, bre- red-shifted in comparison with R. quaesita. This could be the result ams and eels have demonstrated that spectral sensitivities of of an adaptation to the photic environment of coastal waters, which photoreceptors could adjust by expressional patterns of opsin are enriched with higher amounts of particles than the clear coral genes (Carleton, 2009; Carleton & Kocher, 2001; Carleton et al., reef waters. The known ecological information shows that S. sathete 2010; Cottrill et al., 2009; Parry et al., 2005; Shand et al., 2008). is a wide ranging species which can live not only at depths around For the aforementioned arguments, we proposed that the kmax, 300 m, but also in more shallow brackish waters and sometimes 493 and 494 nm, of the green cones of R. quaesita and G. reticularis even ventured into rivers (Myers, 1999; Randall et al., 1990). S. could be achieved by the expression of the Rh2B gene. On the con- sathete displayed a red-shifted of kmax of the cone cells, 509 nm, trary, S. sathete and G. favagineus could express Rh2A in their green which could result from the adaptation to the photic environment cones in which their kmax were longer than 500 nm. Therefore, the of turbid water in estuarine waters (Munz, 1958). spectral shift in the green cones of moray eels could result from the different expression patterns of Rh2 genes. 4.3. Moray eels in this study were all color blinded 4.5. Evolution of opsin genes of moray eels True color vision, i.e. hue discrimination, requires the presence of at least two spectral classes of photoreceptor cell. This is usually In the phylogenetic tree of Rh1 gene (Fig. 4A), the Rh1 genes of accomplished by having two cone classes containing different vi- Anguilliformes were clustered together and formed a monophy- sual pigments (Bowmaker, 1995; Marshall, Vorobyev, & Siebeck, letic group, and that of moray eels were clustered with Rh1 genes 2006). Those species with only one spectral class of cone receptor of freshwater and conger eels. According to the prediction of ances- in their retina are regarded as having ‘‘monochromatic vision’’ at tral sequence of Rh1 gene (Data not shown), the ancestral Rh1 gene diurnal light levels (Bowmaker, 1995). The four moray eel species of Anguilliformes and moray eels both used serine at site 292 of in this study all possessed only one spectral class of green-sensitive Rh1 opsin, the kmax of which was usually around 485 nm (Hunt cone cell. Furthermore, the kmax of rods and cones within species et al., 2001). Further, S292A substitution, that induced a red-shift are similar enough that it is unlikely that the moray eels could of Rh1 opsin, occurred in the lineage of R. quaesita. These results get color vision by rod/cone comparison. These results suggest that implied a possibility that Rh1 genes of moray eels could have all of four moray eels should be considered colorblind. Interest- evolved from deep-sea species first and then later to shallow-sea ingly, the ribbon eel, R. quaesita, is a very colorful moray eel, and species. Gene duplication of Rh1 has occurred before appearance due to its protandrous hermaphrodite nature, exhibits significant of the Anguilliformes, and the deep-sea type Rh1 has been lost in changes of body coloration during sex reversal from black (juve- moray eels. niles, sub-adult), blue (males) to yellow color (females) (Shen, Rh2 gene duplication took place independently several times Lin, & Liu, 1979). Yet, despite these significant changes of body col- and was shown to have evolved independently in cyprinids, cich- oration, the ribbon eel appears colorblind. This seems to imply, lids and moray eels (Fig. 4B). In cichlids, two major groups, Acanth- perhaps, that color vision is not so crucial in the mate recognition opterygii-Rh2A and B, existed and Acanthopterygii-Rh2A could be in a very colorful species like R. quaesita. divided into group 2Aa and 2Ab (Bowmaker, 2008; Parry et al., 2005). In cyprinids, gene duplication occurred before the appear- 4.4. Molecular mechanisms of spectral shift in opsin genes of moray ance of cyprinids and could also be divided into two groups. More- eels over, a similar scenario could also be found during Rh2 genes evolution of seabreams (Wang et al., 2009). In moray eels, gene Amino acid site 292 is one of the spectral tuning sites in Rh1 op- duplication occurred before the appearance of subfamily Muraeni- sins of fish (Hunt, Dulai, Partridge, Cottrill, & Bowmaker, 2001; nae but after the appearance of Anguilliformes. E122Q substitution F.Y. Wang et al. / Vision Research 51 (2011) 1099–1108 1107 has occurred several times and Rh2B of S. sathete has been lost dur- Chave, E. H. N., & Randall, H. A. (1971). Feeding behavior of the moray eel, ing evolutionary history of the Rh2 gene (Fig. 4B). Gymnothorax pictus. Copeia, 1971, 570–574. Chinen, A., Matsumoto, Y., & Kawamura, S. (2005). Reconstitution of ancestral green visual pigments of zebrafish and molecular mechanism of their spectral differentiation. Molecular Biology and Evolution, 22, 1001–1010. 5. Conclusion Collin, S. P. (1997). Specializations of the teleost visual system: Adaptive diversity from shallow-water to deep-sea. Acta Physiologica Scandinavica, 161(Suppl. All four moray eel species studied, namely: R. quaesita, G. favagin- 638), 5–24. Cottrill, P. B., Davies, W. L., Semo, M. a., Bowmaker, J. K., Hunt, D. M., & Jeffery, G. eus, G. reticularis and S. sathete, possess duplex retinas. The morpho- (2009). Developmental dynamics of cone photoreceptors in the eel. BMC logical characteristics suggest that G. favagineus and G. reticularis Developmental Biology, 9, 1–9. could be considered nocturnal species, R. quaesita could be classified Cummings, M. E., & Partridge, J. C. (2001). Visual pigments and optical habitats of surfperch (Embiotocidae) in the California kelp forest. Vision Research, 187, as a diurnal species, and S. sathete could be a crepuscular or diurnal 875–889. species. The MSP data indicated the spectral sensitivities of photo- Davies, W. L., Carvalho, L. S., Tay, B.-H., Brenner, S., Hunt, D. M., & Venkatesh, B. receptors in moray eels were correlated with the photic character- (2009). Into the blue: Gene duplication and loss underlie color vision istics of their habitats. All four moray eel species possessed only adaptations in a deep-sea chimaera, the elephant shark Callorhinchus milii. Genome Research, 19, 415–426. one type of cone cell with sensitivity in the ‘green’ part of the spec- Denton, E. J., & Walker, M. A. (1958). The visual pigment of the conger eel. trum, and hence should be considered colorblind. The spectral shift Proceedings of the Royal Society of London. Series B: Biological Sciences, 148, found in the rod cells of moray eels can be explained by amino acid 257–269. Ebrey, T., & Koutalos, Y. (2001). Vertebrate photoreceptors. Progress in Retinal and substitution at site 292 in the Rh1 gene, while that in cone cells Eye Research, 20, 49–94. could result from differential expression of the Rh2 genes. In sum- Fasick, J. I., & Robinson, P. R. (1998). Mechanism of spectral tuning in the dolphin mary, the results of this study showed that the spectral sensitivities visual pigments. Biochemistry, 37, 433–438. Govardovskii, V. I., Fyhrquist, N., Reuter, T., Kuzmin, D. G., & Donner, a. K. (2000). of moray eels and opsin genes evolved to adapt to different depths In search of the visual pigment template. Visual Neuroscience, 17, 509– or different underwater photic environments. 528. Hasegawa, M., Kishino, H., & Yano, T. (1985). Dating of the human–ape splitting by a molecular clock of mitochondrial DNA. Journal of Molecular Evolution, 22, Acknowledgments 160–174. Hess, M., Melzer, R. R., & Smola, U. (1998). The photoreceptors of Muraena helena and Ariosoma balearicum – A comparison of multiple bank retinae in This work was supported by intramural grants from Academia anguilliform eels (Teleostei). American Zoologist, 237, 127–137. Sinica to H.Y.Y. We thank Dr. Ellis Loew of Cornell University, Dr. Hobson, E. S. (1975). Feeding patterns among tropical reef fishes. American Scientist, Amy Scholik-Schlomer of NMFS, NOAA and Dr. Andy Wigginton 63, 382–392. Hunt, D. M., Dulai, K. S., Partridge, J. C., Cottrill, P., & Bowmaker, J. K. (2001). The of University of Kentucky offered detailed comments on the man- molecular basis for spectral tuning of rod visual pigments in deep-sea fish. uscript and for help with English text editing. Journal of Experimental Biology, 204, 3333–3344. Hunt, D. M., Fitzgibbon, J., Slobodyyanyuk, S. J., & Bowmaker, J. K. (1996). Spectral tuning and molecular evolution of rod visual pigments in the species flock of Appendix A. Supplementary material cottoid fish in Lake Baikal. Vision Research, 36, 1217–1224. Jerlov, N. G. (1968). Optical oceanography. Amsterdam: Elsevier. King, D., & Fraser, V. (2002). More reef fishes and nudibranchs: East and south coast Supplementary data associated with this article can be found, in of Southern Africa. Struik, Cape Town. the online version, at doi:10.1016/j.visres.2011.02.025. Kirk, J. T. O. (1983). Light and photosynthesis in aquatic ecosystems. Cambridge: Cambridge University Press. Kumar, S., Tamura, N., & Nei, M. (2004). MEGA 3: Integrated software for molecular References evolutionary genetics analysis and sequence alignment. Briefings in bioinformatics, 5, 150–163. Lipetz, L. E., & Cronin, T. W. (1988). Application of an invariant spectral form to the Ali, M. A., & Anctil, M. (1976). Retinas of fishes: An atlas. Berlin: Springer-Verlag. visual pigments of crustaceans: Implications regarding the binding of the Archer, S., & Hirano, J. (1996). Absorbance spectra and molecular structure of the chromophore. Vision Research, 28, 1083–1093. blue-sensitive rod visual pigment in the conger eel (Conger conger). Proceedings Loew, E. R. (1994). A third, ultraviolet-sensitive, visual pigment in the Tokay gecko of the Royal Society of London. Series B: Biological Sciences, 263, 761–767. (Gekko gekko). Vision Research, 34, 1427–1431. Archer, S., Hope, A., & Partridge, J. C. (1995). The molecular basis for the green–blue Loew, E. R., & McFarland, W. N. (1990). The underwater visual environment. In R. H. sensitivity shift in the rod visual pigments of the European eel. Proceedings of Douglas & M. Djamgoz (Eds.), The visual system of fish (pp. 1–43). London: the Royal Society of London. Series B: Biological Sciences, 262, 289–295. Chapman and Hall. Bardach, J. E., & Loewenthal, L. A. (1961). Touch receptors in fishes with special Loew, E. R., & Sillman, A. J. (1993). Age-related changes in the visual pigments of the reference to the moray eels (Gymnothorax vicinus and G. Moringa). Copeia, 1961, white sturgeon (Acipenser transmontanus). Canadian Journal of Zoology, 71, 42–46. 1552–1557. Bardach, J. E., Winn, H. E., & Menzel, D. W. (1959). The role of the senses in the Loew, E. R., & Zhang, H. (2006). Propagation of visual signals in the aquatic feeding of the nocturnal reef predators Gymnothorax moringa and G. Vicinus. environment: An interactive windows-based model. In F. Ladich, S. P. Collin, P. Copeia, 1959, 133–139. Moller, & B. G. Kapoor (Eds.). Communication in fishes (Vol. 2, pp. 281–302). Böhlke, E. B., & Randall, J. E. (2000). A review of the moray eels (Angulliformes: Plymouth, England: Science Publishers. Muraenidae) of the Hawaiian Islands, with descriptions of two new species. Losey, G. S., McFarland, W. N., Loew, E. R., Zamzow, J. P., Nelson, P. A., & Marshall, N. Proceedings of the Academy of Natural Sciences of Philadelphia, 150, 203–278. J. (2003). Visual biology of Hawaiian coral reef fishes. I. Ocular transmission and Bowmaker, J. K. (1990). Visual pigments of fishes. In R. H. Douglas & M. Djamgoz visual pigments. Copeia, 2003, 433–454. (Eds.), The visual system of fish (pp. 81–107). London: Chapman and Hall. Marshall, J., Vorobyev, M., & Siebeck, U. E. (2006). What does a reef fish see when it Bowmaker, J. K. (1995). The visual pigments of fish. Progress in Retinal and Eye sees a reef fish? Eating ‘Nemo’. In F. Ladich, S. P. Collin, P. Moller, & B. G. Kapoor Research, 15, 1–31. (Eds.). Communication in fishes (Vol. 2, pp. 393–422). Plymouth, England: Bowmaker, J. K. (2008). Evolution of vertebrate visual pigments. Vision Research, 48, Science Publishers. 2022–2041. McFarland, W. N. (1986). Light in the sea—Correlations with behaviors of fishes and Bowmaker, J. K., & Loew, E. R. (2008). Vision in fish. In R. R. Hoy, G. M. Shepherd, A. I. invertebrates. American Zoologist, 26, 389–401. Basbaum, A. Kaneko, & G. Westheimer (Eds.), The senses: A comprehensive McFarland, W. N. (1991). The visual world of coral reef fishes. In P. F. Sale (Ed.), The reference (Vol. 1) (pp. 53–76). Oxford, UK: Elsevier. ecology of fishes on coral reefs (pp. 16–38). San Diego: Academic Press. Bowmaker, J. K., Semo, M., Hunt, D. M., & Jeffery, G. (2008). Eel visual pigments McFarland, W. N., & Munz, F. W. (1975). Part III: The evolution of photopic visual revisited: The fate of retinal cones during metamorphosis. Visual Neuroscience, pigments in fishes. Vision Research, 15, 1045–1172. 25, 249–255. Morel, A. (1980). In-water and remote measurements of ocean color. Boundary- Carleton, K. (2009). Cichlid fish visual systems: Mechanisms of spectral tuning. Layer Meteorology, 18, 177–201. Integrated Zoology, 4, 75–86. Munz, F. W. (1958). The photosensitive retinal pigments of fishes from relatively Carleton, K. L., Hofmann, C. M., Klisz, C., Patel, Z., Chircus, L. M., Simenauer, L. H., turbid coastal watersrupress. Journal of General Physiology, 42, 445–459. et al. (2010). Genetic basis of differential opsin gene expression in cichlid fishes. Munz, F. W., & McFarland, W. N. (1973). The significance of spectral position in the Journal of Evolutionary Biology, 23, 840–853. rhodopsins of tropical marine fishes. Vision Research, 13, 1829–1874. Carleton, K. L., & Kocher, T. D. (2001). Cone opsin genes of African cichlid fishes: Myers, R. F. (1999). Micronesian reef fishes: A comprehensive guide to the coral reef Tuning spectral sensitivity by differential gene expression. Molecular Biology fishes of Micronesia. Guam: Coral Graphics. and Evolution, 18, 1540–1550. 1108 F.Y. Wang et al. / Vision Research 51 (2011) 1099–1108

Nicol, J. A. C., & Somiya, H. (1989). The eyes of fishes. Oxford: Oxford University Press. Tamura, K., & Nei, M. (1993). Estimation of the number of nucleotide substitutions Ogden, J. C., & Buckman, N. S. (1973). Movements, foraging groups, and diurnal in the control region of mitochondrial DNA in humans and chimpanzees. migrations of the striped parrotfish Scarus croicensis Bloch (Scaridae). Ecology, Molecular Biology and Evolution, 10, 512–526. 54, 589–596. Walls, G. L. (1942). The vertebrate eye and its adaptive radiation. Bloomington Hills, Pankhurst, N. W. (1989). The relationship of ocular morphology to feeding modes Mich., USA: Cranbrook Institute of Science. and activity periods in shallow marine teleosts from New Zealand. Wang, F. Y., Chung, W. S., Yan, H. Y., & Tzeng, C. S. (2008). Adaptive evolution of cone Environmental Biology of Fishes, 26, 201–211. opsin genes in two colorful cyprinids, Opsariichthys pachycephalus and Candidia Parry, J. W. L., Carleton, K. L., Spady, T., Carboo, A., Hunt, D. M., & Bowmaker, J. K. barbatus. Vision Research, 48, 1695–1704. (2005). Mix and match color vision: Tuning spectral sensitivity by differential Wang, F. Y., Yan, H. Y., Chen, J. S.-C., Wang, T. Y., & Wang, D. (2009). Adaptation opsin gene expression in Lake Malawi cichlids. Current Biology, 15, 1734–1739. of visual spectra and opsin genes in seabreams. Vision Research, 49, 1860– Posada, D., & Crandall, K. A. (1998). Modeltest: Testing the model of DNA 1868. substitution. Bioinformatics, 14, 817–818. Warrant, E. J., & Locket, N. A. (2004). Vision in the deep sea. Biological Review, 79, Randall, J. E., Allen, G. R., & Steene, R. C. (1990). Fishes of the Great Barrier Reef and 671–712. Coral Sea. Hawaii: University of Hawaii Press. Winn, H. E., & Bardach, J. E. (1959). Differential food selection by moray eels and a Saitou, L., & Nei, M. (1987). The neighbor-joining method: A new method for possible role of the mucous envelope of parrot fishes in reduction of pewdation. reconstructing phylogenetic trees. Molecular Biology and Evolution, 4, 406–425. Ecology, 40, 296–298. Sandström, A. (1999). Visual ecology of fish – A review with special reference to Yang, Z. (1997). PAML: A program package for phylogenetic analysis by maximum percids. Fiskeriverket Rapport, 2, 45–80. likelihood. Cabios, 13, 555–556. Shand, J., Davies, W. L., Thomas, N., Balmer, L., Cowing, J. A., Pointer, M., et al. (2008). Yang, Z. (2007). PAML 4: Phylogenetic analysis by maximum likelihood. Molecular The influence of ontogeny and light environment on the expression of visual Biology and Evolution, 24, 1586–1591. pigment opsins in the retina of the black bream, Acanthopagrus butcheri. Journal Yokoyama, S. (1994). Gene duplications and evolution of the short wavelength- of Experimental Biology, 211, 1495–1503. sensitive visual pigments in vertebrates. Molecular Biology and Evolution, 11, Shand, J., Hart, N. S., Thomas, N., & Partridge, J. C. (2002). Developmental changes in 32–39. the cone visual pigments of black bream Acanthopagrus butcheri. Journal of Yokoyama, S. (1995). Amino acid replacements and wavelength absorption of visual Experimental Biology, 205, 3661–3667. pigments in vertebrates. Molecular Biology and Evolution, 12, 53–61. Shapley, R., & Gordon, J. (1980). The visual sensitivity of the retina of the conger eel. Yokoyama, S. (1997). Molecular genetic basis of adaptive selection: Examples from Proceedings of the Royal Society of London. Series B: Biological Sciences, 209, color vision in vertebrates. Annual Review of Genetics, 31, 315–336. 317–330. Yokoyama, S. (2000). Molecular evolution of vertebrate visual pigments. Progress in Shen, S. C., Lin, R. P., & Liu, F. C. (1979). Redescription of a protandrous Retinal and Eye Research, 9, 385–419. hermaphroditic moray eel (Rinomuaena quaesita Garmen). Bulletin of the Yokoyama, S. (2008). Evolution of dim-light and color vision pigments. Annual Institute of Zoology Academia Sinica, 18, 79–87. Review of Genomics and Human Genetics, 9, 259–282. Sillman, A. J., Johnson, J. L., & Loew, E. R. (2001). Retinal photoreceptors and visual Yokoyama, S., Tada, T., Zhang, H., & Britt, L. (2008). Elucidation of phenotypic pigments in Boa constrictor imperator. Journal of Experimental Biology, 290, adaptations: Molecular analyses of dim-light vision proteins in vertebrates. 359–365. Proceedings of the National Academy of Sciences of the United States of America, Smith, D. G., & Bohlke, E. B. (1997). A review of the Indo-Pacific banded morays of 105, 13480–13485. the Gymnothorax reticularis group, with descriptions of three new species Yokoyama, S., & Yokoyama, R. (1996). Adaptive evolution of photoreceptors and (Pisces, Anguilliformes, Muraenidae). Proceedings of the Academy of Natural visual pigments in vertebrates. Annual Review of Ecology and Systematics, 27, Sciences of Philadelphia, 148, 177–188. 543–567. Spady, T. C., Seehausen, O., Loew, E. R., Jordan, R. C., Kocher, T. D., & Carleton, K. L. Yokoyama, S., Zhang, H., Radlwimmer, F. B., & Blow, N. S. (1999). Adaptive evolution (2005). Adaptive molecular evolution in the opsin genes of rapidly speciating of color vision of the Comoran coelacanth (Latimeria chalumnae). Proceedings of cichlid species. Molecular Biology and Evolution, 22, 1412–1422. the National Academy of Sciences of the United States of America, 96, 6279–6284. Swofford, D.L. (2000). PAUPÃ. Phylogenetic analysis using parsimony (Ãand other Young, R. F., & Win, H. E. (2003). Activity patterns, diet, and shelter site use for two methods). Sinauer Associates, Sunderland, Massachusetts. species of Moray Eels, Gymnothorax moringa and Gymnothorax vicinus, in Belize. Takenaka, N., & Yokoyama, S. (2007). Mechanisms of spectral tuning in the RH2 Copeia, 2003, 44–55. pigments of Tokay gecko and American chameleon. Gene, 399, 26–32.